ElectionsDU

Summary

An “Election Results Verification” process to read the election resuls reports by UiPath Document Understanding capabilities enhanced with Forms AI extraction method.

✅ Benefit

Offers a fast solution to verification of the results which is crucial during objection period after the elections.

Ensuring a swift and efficient process for verifying election results is of paramount importance, particularly during the period when objections may arise. The need for a streamlined and reliable method for confirming the accuracy and legitimacy of election outcomes cannot be overstated. By providing a rapid solution to this critical aspect of the electoral process, stakeholders can address any discrepancies or concerns in a timely manner, ultimately contributing to the integrity and transparency of the overall electoral system.

UiPath Products

The election process in Türkiye briefly:

The ballot box officers count the votes and create one election result report by hand-filling out a form for every ballot box.
These reports are uploaded to the official sources within a week for political parties and candidates to check, validate, and object to the results.
On the other hand, the Oy ve Ötesi NGO volunteers upload the reports to their platform on the same evening for faster validation. Volunteers read the reports and enter the data manually on the platform.

▶️ My project facilitates this check-validate-object process by pre-screening the reports with the hashtag#UiPath Document Understanding and Apps.

✅ The showcase video demonstrates how high the accuracy of the UiPath Document Understanding handwritten number extraction is.

Process Details

Step 1: Collecting the Documents // UiPath Apps

The process starts with a simple app created with UiPath Apps, which collects the result reports by enabling the capture of the photo.

The user can use the app on their phone to easily capture a photo of the report. The app then uploads the photo of the document to the Orchestrator Storage Bucket.

This part is simulating the similar platform that Oy ve Ötesi has.

Demo of the UiPath Apps. User opens the app, clicks to capture button, camera opens, photo captured and uploaded.

Step 2: Extracting Data // Document Understanding

Uploaded document photos are processed by the pre-trained UiPath Document Understanding ML Skill.

// I trained the ML Skill with as few as 20 documents, and the training was successful enough to detect and extract the hand-writings. //